Advanced Stats Influencing The Ducks and Rest of the League


Sep 25, 2014; Los Angeles, CA, USA; Anaheim Ducks center Andrew Cogliano (7) takes the puck down ice in the first period of the game against the Los Angeles Kings at Staples Center. Mandatory Credit: Jayne Kamin-Oncea-USA TODAY Sports

One of the most groundbreaking movements made this summer and into training camps was the movement of advanced stats being applied more heavily among teams across the NHL. The words “Fancystats, Analytics, Corsi” and “Fenwick” have been a large part of hockey vocabulary this offseason, and a new method of developing and enhancing success has risen.

This method is to base a team’s success and make predictions on said team from data that displays information on puck possession stats. Analytics has become more of a trendy topic and a bit more of a prioritized statistical focal point of teams across the league, after realizing that successful teams like the Kings and Blackhawks track analytics and use them to improve their play. Both Los Angeles and Chicago are notably strong puck possession teams and have enjoyed success winning two Stanley Cups each over the past five seasons.

Four teams all made hires this offseason in the hopes of benefiting their teams based on analytical success.

The New Jersey Devils hired a former poker player, Sunny Mehta, to help apply and implement advanced stats theories into the Devils gameplan and improve their play. The Edmonton Oilers hired Tyler Dellow, a popular hockey blogger known for his knowledge on advanced stats and how they can improve a teams play. The Toronto Maple Leafs, a team whose philosophies had been well-documented to go against the analytics movement, hired a new Assistant GM, Kyle Dubas, who is also known for his background in advanced stats. Toronto also hired Travis Metcalfe, the orchestrator of what was arguably the most popular and most utilized hockey analytics website,, which was taken down by the Maple Leafs. Finally, Eric Tulsky of SB Nation was hired by an unidentified NHL team.

To summarize, NHL teams are starting to pay more attention to advanced stats, and some teams hired people to track advanced metrics for their teams.

I, personally, have never seriously committed to subscribing to the theory of advanced stats and how they truly improve a team, simply because there are scenarios where they aren’t exactly true. Seven of the bottom 15 teams ranked in terms of CF/ 60 minutes were all playoff teams. Like any statistic, Corsi is imperfect because it cannot tell everything, and it has its exceptions. However, advanced statistics show validity because they are good predictors of a team’s ultimate success. LA and Chicago had the two best CF% rates, while Toronto and Buffalo had the two worst. Chicago and Los Angeles are considered to be the cream of the crop in a loaded Western Conference, and they have enjoyed tremendous success over the past few seasons. Their strength in analytics had predicted that they should be upper-echelon teams in the NHL, and they are. Buffalo, on the other hand, finished with the worst record in the NHL last season, while Toronto had a huge downfall the last 14 games of the season last year. With Buffalo, the numbers predicted the Sabres to be one of the worst clubs, which they were. However, Toronto defied what the data was saying because of their record during the first two-thirds of the regular season. However, they collapsed out of a playoff spot because Jonathan Bernier went down with an injury late, and the Maple Leafs’ overwhelming negative shot differential eventually caught up to them.

Yes, analytics in some scenarios can be a good predictor of how good a team is and the direction of where the team is going. However, using advanced stats is not just limited to viewing teams: it can be used to assess players as well.

All the talk of analytics has constantly brought up a debate of how to properly assess a player. Is it done the “old-school way”, with the eye-test, scouting reports, and looking at both raw talent and production from numbers such as goals and assists?  Or is the proper way now to assess a player based on the amount of shot attempts he has compared to the other team and the way he helps improve the teams puck possession stats?

As Ducks fans, we can analyze the individual players on the Anaheim squad. The team leader in terms of Corsi percentage (CF%), or shots on goal + missed shots + blocked shots, was not Ryan Getzlaf or Corey Perry. It was actually Patrick Maroon. Obviously Perry and Getzlaf are two elite players in the NHL and clearly better than Maroon, but Maroon’s possession numbers are numbers that can be better looked at as beneficiary to the team and not to the individual. Numbers like goals and assists are more individually beneficial. Obviously, scoring a goal is good for the team. It gets the team on the scoreboard: everyone knows that. However, the thought process behind Corsi and Fenwick (same as Corsi, except it doesn’t count blocked shots) is that the longer a team (or player) has the puck, the more shots they will be able to take. With more shots, theoretically, should come more goals. Essentially, when non-elite and bottom-six players have better Corsi and Fenwick numbers than the elite players on a team, it doesn’t mean that they are significantly better. It essentially means that they contribute to the team in a different way.

Also, as elite players, Getzlaf and Perry face tougher defensive assignments than a player like Maroon will. Corsi and Fenwick can be expressed as individual figures, but the number is largely influenced by both the quality of opponents the individual player faces and the quality of linemates that a player has. Maroon played a lot of minutes with now-departed center Mathieu Perreault, and they faced relatively easier opponents than Getzlaf and Perry. Perreault was also the Ducks’ best possession centerman, and has been strong in that department throughout his career. The combination of playing easier opponents with a center who usually drives play helps boost Maroon’s numbers to the top spot. This is why ExtraSkater made a large point in adding data of Quality of Competition and Quality of Teammates as well.

Are the Ducks’ possession numbers so bad that they need to bring in experts in the field of advanced statistics? No. The Ducks were ranked 15th in the league in CF/60 and ranked 12th in the league in FF/60. The only negative aspect of the Ducks, in terms of advanced statistics, is that they lost two players who had strong possession numbers (Stephane Robidas and Perreault) while signing Clayton Stoner, who had just a 47.1 CF% and Dany Heatley, who had a CF% of just 44.5%. Obviously they brought in Heatley to add scoring depth, potentially on the top line, while signing Stoner to add a more physical presence to their defense, so analytics probably did not play a major factor in their respective signings.

So will the advanced stats wave ever make its way over to Anaheim? Honestly, the Ducks may not need to go add analytics people. Yes, the Ducks signed two players who aren’t great in terms of advanced statistics. However, they also traded for Ryan Kesler, a player who is strong in possession numbers: he had the 6th highest CF% in the entire NHL in 2010-2011, when he won the Selke Trophy, and his CF% last season was higher than that of any center on the Ducks in 2013-2014. The young players, such as Sami Vatanen and Hampus Lindholm, are the players who generally have better possession numbers compared to the possession numbers of other players on the team (however, remember the quality of competition and linemates factor). Right now, it’s not a necessity for the Ducks to look deep into their staff and try to thoroughly assess their players based on analytics and  improve on their advanced stats. This is a team that is explosive on offense. There are two ways to be successful despite bad possession numbers: supreme goaltending (to offset the large volume of shots against) and supreme finishers at the net (to compensate for the small volume of shots for). The Ducks had the best team shooting percentage in the league last year at 5-on-5 play, and although that figure should come down, they still have weapons such as Perry, Getzlaf, and Kesler. Despite them not being a top-end possession club, the Ducks should be a contender in the Western Conference and could be playing for the Stanley Cup in June.

The Ducks also probably have access to this data. Think about what exactly Corsi and Fenwick say: more puck possession leads to more shots, which should lead to more goals. Usually, elite players have the puck on their stick a lot more than they don’t, hence they generally have better puck possession figures. So “advanced stats” and conventional hockey wisdom are not that far apart: they just express things differently: one numerically and analytically, and the other descriptively and visually. Honestly, thinking about advanced stats is probably the last thing on the minds of the Ducks’ executives.

At this point, the only thing the Ducks’ front office should be caring about is trying to win a Stanley Cup.